This study analyzes methods to structure and visualize bibliographic recommendations efficiently while conveying important information for users. Most of the work being developed in the realm of bibliography visualizations is surrounding the question of how authors are related and the network concerning their progression throughout their career. The aspect of conveying information about papers and why to read them have become or are the secondary idea when analyzing these graphs. Why should a user choose a paper over another when a visualization technique or algorithm can aid in the decision-making process? Our visualization should be the reverse/opposite of past work; Co-authorship networks are secondary with a primary emphasis on the network of recommended papers that the user wants to read. This study provides a structured pipeline for better viewing bibliographic recommendations and their relations. This method makes use of important machine learning techniques such as word embeddings and self-organizing maps to take extracted topics/key phrases and map relations to other recommended papers. This extends the typical node-link graph with links representing relations and provides spatial relations of papers that are more intuitive to a user. This study also provides exposure of metadata for customizable aspects of the visualizations for interactive searching. In addition to a 2D view of the recommendations, a side-by-side 3D view is provided for quantitative values.